What documented workflows can I build?
You can build webhook automation that receives an external event, processes data, optionally applies AI, and finishes with an action or response. This hub links 3 documented paths: webhook and AI automation, sending structured results to Google Sheets, and capturing website leads into Google Sheets. Choose by your input and intended result, then check connections, configuration, and credentials.
By Ai Flowi · Technical review: Janice Lee, Technical Support, Ai Flowi.
Published · Updated
On 1 October 2026, an editor test with a sample webhook body wrote one row to an app-created Google Sheet through Universal Webhook → JSON Operations (JQ) → Append Rows, with no AI model. Production-endpoint delivery, sheets not created by the app, and field validation were not exercised. See AI Agent setup and training.
Public use cases
- Webhook and AI automation explains when to receive an event through Universal Webhook, use ordinary rules or processing, and add AI when the input calls for it.
- Webhook to Google Sheets tutorial follows a documented path from webhook JSON through type conversion and Structured Output to Google Sheets Append Rows.
- Capture website leads into Google Sheets explains how a website submission becomes a validated JSON request and a spreadsheet row.
Use the workflow template library to find a task-specific starting point, then review its configuration and test scope.
To validate required webhook fields, inspect Webhook input validator and review the input, output and configuration before using it.
To prepare lead data before the write, inspect Clean lead data and review the input, output and configuration before using it.
Where should I start with a documented workflow?
If your source system sends an event, start with the webhook pattern and identify the request data you need. If the result belongs in a sheet, check the Google Sheets path and its required connection. Inspect input and output types before connecting components, then test configuration and credentials. See demo booking.
The webhook to Google Sheets tutorial gives a detailed build path. For a smaller first exercise, use your first workflow. See AI Flowi Workflow.
Use workflow planning tools to prepare synthetic inputs before configuring a starting pattern.
What are examples of AI workflow?
Examples include Document Q&A, which answers questions about a PDF, Meeting Summary, which transcribes audio and generates a summary, and Memory Chatbot, which saves and references previous messages. These starter templates provide configured starting points. Review the components, supply required credentials and data, then test the workflow before using its output.
Documents and data
Document Q&A
Reads a PDF and uses a language model to answer questions about its contents. Needs a model connection, a readable PDF and a question about the document.
Financial Report Parser
Extracts financial metrics such as Gross Profit, EBITDA and Net Income into structured output. Needs a model key and financial report text for the Structured Output component.
Hybrid Search RAG
Explores hybrid search with a vector database. Needs a model key, an Astra DB connection and a prepared collection of documents to search. For that retrieval design, follow the hybrid search RAG guide before preparing the document collection.
Invoice Summarizer
Uses Needle retrieval with an agent for invoice-related work. Needs a model key, a Needle API key, a collection identifier and invoice data in that collection.
Knowledge Base
Runs a vector search against a Knowledge Base to retrieve relevant documents. Needs a selected Knowledge Base with searchable data and a search query; this retrieval template has no separate chat-model component.
Meeting Summary
Transcribes meeting audio and generates a summary using AssemblyAI and OpenAI. Needs transcription credentials, a model key and an accessible meeting audio file or URL. To plan the document after the summary, explore meeting minutes in Google Docs as the next workflow task.
Text Sentiment Analysis
Loads text from files and classifies sentiment through model prompts. Needs a model key and a readable text-data file.
Vector Store RAG
Loads data to provide chat context through retrieval augmented generation. Needs a model connection, source files and a selected Knowledge Base for ingestion and retrieval.
Agents and research
Market Research
Researches companies and turns business information into structured output. Needs a model key, a Tavily search key and the company or research question.
News Aggregator
Extracts information from webpages. Needs a model key, an AgentQL key, a target webpage and a writable output location.
Pokédex Agent
Uses a specialised agent and the Pokédex API to research Pokémon. Needs a model key, a configured API request and a Pokémon question.
Price Deal Finder
Searches and compares product prices across e-commerce platforms. Needs a model key, Tavily and AgentQL credentials, and a product search request.
Research Agent
Generates a research plan, searches the web and combines findings into a report. Needs model access, a Tavily search key and a research question.
Research Translation Loop
Loops over search results and translates each result into Portuguese. Needs a model key and an arXiv search query.
Search agent
Searches the web for information. Needs a model key, a ScrapeGraph search API key and a search question.
Sequential Tasks Agents
Runs agent tasks in a predefined sequence. Needs model access, a Tavily search key and task input; its tools include financial data retrieval and a calculator.
Simple Agent
Provides a starter agent workflow. Needs a model key and a chat request; its URL and calculator tools also need suitable task input.
Travel Planning Agents
Uses specialised agents to produce trip itineraries in a travel-planning chat. Needs model keys, a search API key and travel requirements.
Content and marketing
Blog Writer
Drafts a blog post from instructions and referenced articles. Needs a model key, writing instructions and accessible reference URLs.
Instagram Copywriter
Creates Instagram post copy and image prompts. Needs model access, a Tavily search key and the post topic or content brief.
SEO Keyword Generator
Produces keyword ideas from product information, pain points and customer profiles. Needs a model key and those product and audience details.
Social Media Agent
Uses Apify Actors as tools to search and analyse social media profiles. Needs a model key, an Apify token and the profile or search input required by the selected Actors.
Twitter Thread Generator
Turns structured inputs into a Twitter thread through a prompt-based workflow. Needs a model key, profile details, content guidelines, output format, language and tone settings.
YouTube Analysis
Extracts video comments and transcripts to examine sentiment and content themes. Needs a model key, a video URL and the API key used by the YouTube Comments component.
Chat and memory
Basic Prompt Chaining
Connects prompts in sequence so each output feeds the next stage. Needs model access, prompt instructions and a text input.
Basic Prompting
Runs basic prompting with an OpenAI model. Needs a model key, a prompt and chat input.
Memory Chatbot
Saves and references previous messages to provide conversation context. Needs a model key, chat input and configured message history. Read the agent memory overview when deciding how previous messages should inform the next response.
Other
Custom Component Generator
Generates custom-component code following AI Flowi Workflow specifications. Needs a model key, a component request and accessible specification URLs; review generated code before using it.
Image Sentiment Analysis
Uses zero-shot classification to categorise images as positive, negative or neutral. Needs model access and image-related input; test the selected model and input format before using the result.
NVIDIA RTX Remix
Connects an agent to the NVIDIA RTX Remix Toolkit API and its documentation. Needs a model key, embedding-model access, a configured MCP connection and accessible RTX Remix documentation.
Portfolio Website Code Generator
Converts a PDF or TXT resume into structured JSON and then portfolio website HTML code. Needs model access and a readable resume file; review the generated code before using it.
SaaS Pricing
Calculates a SaaS subscription price from costs, profit margin and subscriber input. Needs a model key and the cost, margin and subscriber inputs for the calculation; it does not describe Ai Flowi subscription pricing.